[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: POLARIZATION+CFAVERAGE
Dear Maurizio,
I attached few comments to your letter. Please, send me your feff.inp
file, so it will be easier for me to debug the problem.
Best wishes
Alex Ankudinov
On Tue, 8 Oct 2002, Maurizio Mattesini wrote:
>
> I'd like to ask you few questions about the correct use of the POLARIZATION
> and the CFAVERAGE cards. In particular, how should one employ the
> POLARIZATION card in systems where more than one atom sites are present?
> Should one calculate the XANES spectra by first running feff8.20 with the
> CFAVERAGE card and then plug the POLARIZATION card? or is it possible to
The CFAVERAGE card was designed for EXAFS calculations. I.e. it assumes
phase transferability for atoms at different sites. While this
approximation is well justified for EXAFS region, it might fail for XANES.
For XANES region I usually do not use CFAVERAGE card, but calculate
absorption for inequivalent sites, and then sum them up with appropriate
weights. I never checked whether I would get the same result with
CFAVERAGE card.
> specify the two cards at the same time in the datafile? Which is the correct
> procedure to follow in order to get a XANES spectra that accounts for the
> configurational average around the different atom sites and the effects of
> the polarization?
It is O'k to use XANES and POLARIZATION in the same input file,
but as I said above the use of CFAVERAGE and XANES has never been
seriously tested.
> One last question: I tried to use the CFAVERAGE without the potential index 0
> in the input file but I always got the same errorr (No atoms or overlap cards
> for unique pot 0 Cannot calculate potentials). Please note that the
> user guide clearly says that it it not necessary to have the potential index
> 0 when using the CFAVERAGE card (top of p.10 of the manual). On the other
This sound as a bug. I need feff.inp file to address this question.
> hand when specifying the configurational average for all the atoms with
> potential index 2 (i.e. CFAVERAGE 2 0 0) the calculation ends without
> writing the xmu.dat file. I guess I'm doing something wrong, any comments or
> suggestions will be highly appreciated!!
This might be connected with the fact that CFAVERAGE was designed for
EXAFS, and not for XANES. Did you get chi.dat file? This also sounds as
a bug and it will be better if I use your feff.inp file.
I would suggest not to use CFAVERAGE and XANES cards together, until
it has been tested. Thus to get average absorption, I would get
xmu.dat file for each absorption sight and average them.
Alex Ankudinov